@measureoneinc/savant 1.0.0

package/dist/services/embedding.service.js

@@ -0,0 +1,587 @@
+"use strict";
+/**
+ * @fileoverview Embedding Service
+ *
+ * Service for generating embeddings using Gemini and storing them in the database.
+ *
+ * @author MeasureOne, Inc.
+ * @version 1.0.0
+ */
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.EmbeddingService = void 0;
+const genai_1 = require("@google/genai");
+const logger_1 = require("../lib/utils/logger");
+const models_1 = require("../models");
+const uuid_1 = require("uuid");
+const config_1 = require("../config");
+const sequelize_service_1 = require("../lib/sequelize-service");
+const sequelize_1 = require("sequelize");
+const faiss_node_1 = require("faiss-node");
+const logger = (0, logger_1.create_default_logger)("embedding-service");
+/**
+ * Chunk text into smaller pieces with overlap
+ */
+function create_chunks(text, chunk_size = 1200, overlap = 150) {
+    const chunks = [];
+    let start = 0;
+    while (start < text.length) {
+        const end = Math.min(start + chunk_size, text.length);
+        chunks.push(text.slice(start, end));
+        if (end >= text.length)
+            break;
+        start = Math.max(start + 1, end - overlap);
+    }
+    return chunks;
+}
+class EmbeddingService {
+    constructor() {
+        this.default_model = "text-embedding-004"; // Gemini embedding model
+        // Query embedding cache (Phase 1 optimization)
+        // Primary strategy: Semantic similarity (embeddings indexed by hash)
+        // Secondary strategy: Exact text match (fast path, but rare for user queries)
+        this.query_embedding_cache = new Map();
+        this.EMBEDDING_CACHE_TTL_MS = 60 * 60 * 1000; // 1 hour
+        this.MAX_EMBEDDING_CACHE_SIZE = 1000;
+        this.SEMANTIC_SIMILARITY_THRESHOLD = 0.92; // Lower threshold for better cache hits (was 0.95)
+        this.MAX_SEMANTIC_CHECKS = 50; // Limit similarity checks to prevent O(n) on large cache
+        // In-memory Faiss index for fast vector search
+        this.faiss_index_cache = new Map();
+        this.is_index_building = false;
+        const api_key = config_1.config_utils.get("GEMINI_API_KEY");
+        if (!api_key) {
+            throw new Error("GEMINI_API_KEY not configured");
+        }
+        this.gemini = new genai_1.GoogleGenAI({ apiKey: api_key });
+    }
+    /**
+     * Builds or retrieves a cached Faiss index for a given model.
+     * This is the core of the search optimization.
+     * @param model - The embedding model name (e.g., "text-embedding-004")
+     */
+    async get_or_build_faiss_index(model) {
+        if (this.faiss_index_cache.has(model)) {
+            return this.faiss_index_cache.get(model);
+        }
+        if (this.is_index_building) {
+            logger.info("Faiss index is already being built, waiting...");
+            // Simple wait loop, can be improved with event emitters for production
+            await new Promise(resolve => setTimeout(resolve, 2000));
+            return this.get_or_build_faiss_index(model);
+        }
+        try {
+            this.is_index_building = true;
+            logger.info(`Building Faiss index for model: ${model}...`);
+            const embeddings = await models_1.EmbeddingModel.findAll({
+                where: { model },
+                attributes: ['id', 'embedding', 'text', 'knowledge_id', 'knowledge_revision_id', 'source_data_id', 'chunk_index', 'type', 'domain'],
+            });
+            if (embeddings.length === 0) {
+                logger.warn(`No embeddings found for model ${model} to build index.`);
+                return null;
+            }
+            const dimension = JSON.parse(embeddings[0].embedding).length;
+            const index = new faiss_node_1.Index(dimension);
+            const vectors = [];
+            const metadata = [];
+            for (const emb of embeddings) {
+                try {
+                    const vector = JSON.parse(emb.embedding);
+                    // Validate vector dimension
+                    if (!Array.isArray(vector) || vector.length !== dimension) {
+                        logger.warn(`Skipping embedding with invalid dimension: ${emb.id} (expected ${dimension}, got ${Array.isArray(vector) ? vector.length : 'not an array'})`);
+                        continue;
+                    }
+                    vectors.push(vector);
+                    metadata.push({
+                        id: emb.id,
+                        text: emb.text,
+                        knowledge_id: emb.knowledge_id,
+                        knowledge_revision_id: emb.knowledge_revision_id,
+                        source_data_id: emb.source_data_id,
+                        chunk_index: emb.chunk_index,
+                        type: emb.type,
+                        domain: emb.domain,
+                    });
+                }
+                catch (e) {
+                    logger.warn(`Skipping invalid embedding record: ${emb.id}`, { error: e instanceof Error ? e.message : String(e) });
+                }
+            }
+            if (vectors.length > 0) {
+                // Add vectors one at a time (faiss-node add method expects single vector)
+                for (const vector of vectors) {
+                    index.add(vector);
+                }
+                const index_data = { index, metadata };
+                this.faiss_index_cache.set(model, index_data);
+                logger.info(`✅ Faiss index built for model ${model} with ${index.ntotal()} vectors.`);
+                return index_data;
+            }
+            return null;
+        }
+        catch (error) {
+            logger.error(`❌ Failed to build Faiss index for model ${model}`, { error });
+            return null;
+        }
+        finally {
+            this.is_index_building = false;
+        }
+    }
+    /**
+     * Generate embedding with caching (Phase 1 optimization)
+     * Caches query embeddings to avoid redundant API calls
+     *
+     * Cache strategy (optimized for user-generated queries):
+     * 1. Quick exact text check (O(1) - fast path, but rare for user queries)
+     * 2. Generate embedding and check semantic similarity (primary strategy)
+     * 3. Cache by semantic similarity, not exact text (since exact matches are rare)
+     *
+     * Rationale: User queries are semantically similar but textually different.
+     * Exact text caching has low hit rate, so we prioritize semantic similarity.
+     */
+    async generate_embedding_cached(text, model) {
+        const normalized_text = text.toLowerCase().trim();
+        const exact_cache_key = `${model}:${normalized_text}`;
+        // Step 1: Quick exact text check (O(1) - worth checking but expect low hit rate)
+        const exact_cached = this.query_embedding_cache.get(exact_cache_key);
+        if (exact_cached && Date.now() - exact_cached.timestamp < this.EMBEDDING_CACHE_TTL_MS) {
+            logger.debug('Query embedding cache hit (exact match)', {
+                cache_key: exact_cache_key.substring(0, 50) + '...',
+                cache_size: this.query_embedding_cache.size
+            });
+            return exact_cached.embedding;
+        }
+        // Step 2: Generate embedding (we need it anyway for semantic comparison)
+        const query_embedding = await this.generate_embedding(text, model);
+        const query_norm = Math.sqrt(query_embedding.reduce((sum, val) => sum + val * val, 0));
+        // Step 3: Check for semantically similar cached embeddings (primary strategy)
+        // Only check if cache has entries
+        if (this.query_embedding_cache.size > 0) {
+            const similar_match = this.find_similar_cached_embedding_sync(query_embedding, query_norm, model, this.SEMANTIC_SIMILARITY_THRESHOLD);
+            if (similar_match) {
+                logger.debug('Query embedding cache hit (semantic match)', {
+                    query_text: text.substring(0, 50) + '...',
+                    cache_size: this.query_embedding_cache.size,
+                    threshold: this.SEMANTIC_SIMILARITY_THRESHOLD
+                });
+                // Cache exact text too (in case same user retries exact query)
+                // But don't duplicate the embedding - just reference it
+                this.set_cache_entry(exact_cache_key, query_embedding, query_norm, text);
+                return similar_match;
+            }
+        }
+        // Step 4: No match found - cache the new embedding
+        logger.debug('Query embedding cache miss, caching new embedding', {
+            cache_key: exact_cache_key.substring(0, 50) + '...',
+            cache_size: this.query_embedding_cache.size
+        });
+        this.set_cache_entry(exact_cache_key, query_embedding, query_norm, text);
+        return query_embedding;
+    }
+    /**
+     * Set cache entry with LRU eviction
+     * Stores embedding with pre-computed norm for faster similarity calculations
+     */
+    set_cache_entry(cache_key, embedding, embedding_norm, original_text) {
+        // Pre-compute norm if not provided
+        const norm = embedding_norm || Math.sqrt(embedding.reduce((sum, val) => sum + val * val, 0));
+        // LRU eviction: Remove oldest entry if cache is full
+        if (this.query_embedding_cache.size >= this.MAX_EMBEDDING_CACHE_SIZE) {
+            const oldest_key = Array.from(this.query_embedding_cache.entries())
+                .sort((a, b) => a[1].timestamp - b[1].timestamp)[0][0];
+            this.query_embedding_cache.delete(oldest_key);
+            logger.debug('Evicted oldest cache entry', {
+                evicted_key: oldest_key.substring(0, 50) + '...',
+                remaining_size: this.query_embedding_cache.size
+            });
+        }
+        this.query_embedding_cache.set(cache_key, {
+            embedding,
+            embedding_norm: norm,
+            timestamp: Date.now(),
+            original_text: original_text || cache_key
+        });
+    }
+    /**
+     * Find semantically similar cached embedding (synchronous version)
+     * Checks cached embeddings for high similarity (>= threshold)
+     *
+     * Optimization: Limits checks to prevent O(n) performance on large cache
+     * Uses pre-computed norms for faster similarity calculation
+     *
+     * @param query_embedding - Pre-computed query embedding vector
+     * @param query_norm - Pre-computed query embedding norm
+     * @param model - Embedding model
+     * @param similarity_threshold - Minimum similarity threshold
+     * @returns Matching embedding or null
+     */
+    find_similar_cached_embedding_sync(query_embedding, query_norm, model, similarity_threshold) {
+        const now = Date.now();
+        let checked_count = 0;
+        let best_match = null;
+        // Check cached embeddings (limit checks for performance)
+        for (const [cache_key, cached] of this.query_embedding_cache.entries()) {
+            // Only check embeddings from the same model
+            if (!cache_key.startsWith(`${model}:`)) {
+                continue;
+            }
+            // Skip if cache entry is expired
+            if (now - cached.timestamp >= this.EMBEDDING_CACHE_TTL_MS) {
+                continue;
+            }
+            // Limit number of similarity checks to prevent O(n) on large cache
+            if (checked_count >= this.MAX_SEMANTIC_CHECKS) {
+                break;
+            }
+            // Calculate similarity using pre-computed norms
+            const similarity = this.cosine_similarity_optimized(query_embedding, cached.embedding, query_norm, cached.embedding_norm);
+            checked_count++;
+            // Track best match (even if below threshold, in case we want to lower threshold later)
+            if (!best_match || similarity > best_match.similarity) {
+                best_match = { embedding: cached.embedding, similarity };
+            }
+            // Return immediately if we find a match above threshold
+            if (similarity >= similarity_threshold) {
+                logger.debug('Found semantically similar cached embedding', {
+                    cache_key: cache_key.substring(0, 50) + '...',
+                    similarity: similarity.toFixed(4),
+                    checked_count
+                });
+                return cached.embedding;
+            }
+        }
+        // Log if we checked but didn't find a match
+        if (checked_count > 0 && best_match) {
+            logger.debug('No semantic match found above threshold', {
+                best_similarity: best_match.similarity.toFixed(4),
+                threshold: similarity_threshold,
+                checked_count,
+                cache_size: this.query_embedding_cache.size
+            });
+        }
+        return null;
+    }
+    /**
+     * Generate embedding for a single text using Gemini
+     *
+     * @param text - Text to embed
+     * @param model - Embedding model (default: text-embedding-004)
+     * @returns Promise resolving to embedding vector
+     */
+    async generate_embedding(text, model) {
+        try {
+            const embedding_model = model || this.default_model;
+            logger.debug("Generating embedding using Gemini", {
+                model: embedding_model,
+                text_length: text.length,
+            });
+            // Use Gemini's embedding API
+            // The @google/genai package uses models.embedContent method
+            const result = await this.gemini.models.embedContent({
+                model: embedding_model,
+                contents: [{
+                        parts: [{ text }]
+                    }]
+            });
+            // Extract embedding values from the response
+            // Gemini returns embeddings array, get the first one
+            const embedding_values = result.embeddings?.[0]?.values;
+            if (!embedding_values || embedding_values.length === 0) {
+                throw new Error("No embedding returned from Gemini");
+            }
+            logger.debug("Embedding generated successfully", {
+                model: embedding_model,
+                embedding_dimension: embedding_values.length,
+            });
+            return embedding_values;
+        }
+        catch (error) {
+            logger.error("Failed to generate embedding", { error });
+            throw new Error(`Failed to generate embedding: ${error instanceof Error ? error.message : "Unknown error"}`);
+        }
+    }
+    /**
+     * Create embeddings for a document (works for both SEED and LEARNING types)
+     * This method chunks text, generates embeddings, and stores them in the database.
+     *
+     * @param knowledge_id - Required knowledge ID (must exist)
+     * @param knowledge_revision_id - Required revision ID (must exist)
+     * @param text - Text content to create embeddings for
+     * @param source_data_id - Optional source data ID
+     * @param model - Optional embedding model (defaults to default_model)
+     * @param type - Embedding type (SEED or LEARNING)
+     * @param domain - Required domain (normalized to lowercase)
+     *
+     * Note: This does NOT search for similar documents or create knowledge/revision.
+     * The caller must ensure knowledge_id and knowledge_revision_id already exist.
+     */
+    async create_embeddings_for_document(knowledge_id, knowledge_revision_id, text, source_data_id, model, type, domain) {
+        // Validate required parameters
+        if (!knowledge_id || !knowledge_revision_id) {
+            throw new Error("knowledge_id and knowledge_revision_id are required for all embedding types");
+        }
+        // Validate domain
+        if (!domain || domain.trim().length === 0) {
+            throw new Error("Domain is required for all embedding types (SEED and LEARNING)");
+        }
+        // Normalize domain to lowercase for consistency
+        const normalized_domain = domain.trim().toLowerCase();
+        const embedding_model = model || this.default_model;
+        logger.info("Processing document for embeddings", {
+            knowledge_id,
+            knowledge_revision_id,
+            chunks_count: 0, // Will be set after chunking
+            model: embedding_model,
+            domain: normalized_domain
+        });
+        // Chunk the text
+        const chunks = create_chunks(text, 1200, 150);
+        logger.info("Document chunked", {
+            knowledge_id,
+            chunks_count: chunks.length,
+            domain: normalized_domain
+        });
+        // STEP 1: Generate all embeddings FIRST (outside transaction)
+        // This is the slow part (API calls) and should NOT hold a database lock
+        const all_embedding_records = [];
+        const batch_size = 10;
+        for (let i = 0; i < chunks.length; i += batch_size) {
+            const batch = chunks.slice(i, i + batch_size);
+            // Generate embeddings for batch (API calls - can take seconds/minutes)
+            const embedding_promises = batch.map((chunk, index) => this.generate_embedding(chunk, embedding_model).then((embedding) => ({
+                chunk_index: i + index,
+                text: chunk,
+                embedding,
+            })));
+            const results = await Promise.all(embedding_promises);
+            // Prepare embedding records for batch insert
+            for (const result of results) {
+                const embedding_record = {
+                    id: (0, uuid_1.v4)(),
+                    knowledge_id,
+                    knowledge_revision_id,
+                    source_data_id,
+                    text: result.text,
+                    embedding: JSON.stringify(result.embedding),
+                    model: embedding_model,
+                    chunk_index: result.chunk_index,
+                    type: type, // Use the type parameter (SEED or LEARNING)
+                };
+                // Domain is required and guaranteed to be valid at this point
+                // Use normalized_domain to ensure lowercase storage
+                embedding_record.domain = normalized_domain;
+                all_embedding_records.push(embedding_record);
+            }
+            logger.debug('Prepared embedding records', {
+                total_records: all_embedding_records.length,
+                domain
+            });
+        }
+        // STEP 2: Only NOW start transaction for database writes
+        // Transaction is held open only for the brief database write operation
+        const sequelize = (0, sequelize_service_1.get_sequelize_service)();
+        const transaction = await sequelize.transaction();
+        try {
+            let embeddings_created = 0;
+            // Batch insert all embeddings in a single transaction (fast operation)
+            if (all_embedding_records.length > 0) {
+                await models_1.EmbeddingModel.bulkCreate(all_embedding_records, { transaction });
+                embeddings_created = all_embedding_records.length;
+            }
+            await transaction.commit();
+            logger.info("Document processed successfully", {
+                knowledge_id,
+                knowledge_revision_id,
+                embeddings_created,
+            });
+            return { embeddings_created, knowledge_id, knowledge_revision_id };
+        }
+        catch (error) {
+            await transaction.rollback();
+            logger.error("Failed to process document", { knowledge_id, knowledge_revision_id, error });
+            throw error;
+        }
+    }
+    /**
+     * Find knowledge ID for learned patterns
+     */
+    async find_learned_patterns_knowledge_id() {
+        try {
+            const app_id = config_1.config_utils.get("APPLICATION_ID") || "savant-orchestration";
+            const requester_id = config_1.config_utils.get("REQUESTER_ID") || "execution-learner";
+            const knowledge_records = await models_1.KnowledgeModel.findAll({
+                where: {
+                    application_id: app_id,
+                    requester_id: requester_id,
+                    metadata: {
+                        [sequelize_1.Op.like]: '%"purpose"%learned-patterns%'
+                    }
+                },
+                limit: 1,
+                order: [["created_at", "DESC"]]
+            });
+            return knowledge_records.length > 0 ? knowledge_records[0].id : null;
+        }
+        catch (error) {
+            logger.debug("Could not find learned patterns KB", { error });
+            return null;
+        }
+    }
+    /**
+     * Search embeddings by similarity using a two-pass top-k approach
+     *
+     * @param query_text - Query text to search for
+     * @param knowledge_id - Optional knowledge ID to filter by (learned knowledge from LEARNING embeddings is ALWAYS included)
+     * @param limit - Maximum number of results (default: 10)
+     * @param model - Embedding model (default: text-embedding-004)
+     * @returns Promise resolving to array of similar embeddings with scores
+     *
+     * This optimized version uses a pre-built in-memory Faiss index for fast vector search.
+     */
+    async search_embeddings(query_text, knowledge_id, limit = 10, model, type, domain) {
+        try {
+            const embedding_model = model || this.default_model;
+            // If filters are provided, query database first to get matching embedding IDs
+            // This allows us to filter Faiss results more efficiently
+            let allowed_embedding_ids = null;
+            if (knowledge_id || type || domain) {
+                const where_clause = { model: embedding_model };
+                if (knowledge_id) {
+                    where_clause.knowledge_id = knowledge_id;
+                }
+                if (type) {
+                    where_clause.type = type;
+                }
+                if (domain) {
+                    // Domains are stored in lowercase, so we can use simple equality
+                    where_clause.domain = domain.trim().toLowerCase();
+                }
+                const matching_embeddings = await models_1.EmbeddingModel.findAll({
+                    where: where_clause,
+                    attributes: ['id'],
+                });
+                if (matching_embeddings.length === 0) {
+                    logger.debug("No embeddings found matching filters", { knowledge_id, type, domain });
+                    return [];
+                }
+                allowed_embedding_ids = new Set(matching_embeddings.map(e => e.id));
+                logger.debug("Filtered embeddings by criteria", {
+                    matching_count: allowed_embedding_ids.size,
+                    knowledge_id,
+                    type,
+                    domain
+                });
+            }
+            // Build or get the cached Faiss index for the model.
+            const index_data = await this.get_or_build_faiss_index(embedding_model);
+            if (!index_data) {
+                logger.warn(`Faiss index not available for model ${embedding_model}. Cannot perform search.`);
+                return [];
+            }
+            // Generate embedding for query (using cached version)
+            const query_embedding = await this.generate_embedding_cached(query_text, embedding_model);
+            // Find learned patterns KB
+            const learned_patterns_knowledge_id = await this.find_learned_patterns_knowledge_id();
+            // Perform the search against the Faiss index
+            // Search for more results if we have filters (to account for filtering)
+            // But ensure we don't exceed the total number of vectors in the index
+            const index_ntotal = index_data.index.ntotal();
+            const requested_limit = allowed_embedding_ids ? limit * 3 : limit * 2;
+            const search_limit = Math.min(requested_limit, index_ntotal);
+            if (search_limit === 0) {
+                logger.debug("Index is empty, returning empty results");
+                return [];
+            }
+            const { labels, distances } = index_data.index.search(query_embedding, search_limit);
+            // Map results back to our metadata and filter by allowed IDs if filters were provided
+            const all_results = labels
+                .map((label, i) => {
+                const metadata = index_data.metadata[label];
+                // Faiss returns distance, convert to similarity (1 - distance for cosine)
+                const score = 1 - distances[i];
+                return { ...metadata, score };
+            })
+                .filter((result) => {
+                // If we have a filter set, only include results whose embedding ID is in the allowed set
+                if (allowed_embedding_ids && !allowed_embedding_ids.has(result.id)) {
+                    return false;
+                }
+                return true;
+            });
+            // Sort by score and return top-k
+            const final_results = all_results
+                .sort((a, b) => b.score - a.score)
+                .slice(0, limit);
+            if (final_results.length === 0) {
+                logger.warn("No similar embeddings found in Faiss index", { knowledge_id, type, domain });
+                return [];
+            }
+            logger.debug("Faiss index search completed", {
+                initial_found: labels.length,
+                filtered_count: all_results.length,
+                final_count: final_results.length,
+                limit,
+                had_filters: !!allowed_embedding_ids
+            });
+            return final_results;
+        }
+        catch (error) {
+            logger.error("Failed to search embeddings", { error });
+            throw error;
+        }
+    }
+    /**
+     * Calculate cosine similarity between two vectors
+     *
+     * @param vec_a - First vector
+     * @param vec_b - Second vector
+     * @returns Cosine similarity score (0-1)
+     */
+    cosine_similarity(vec_a, vec_b) {
+        if (vec_a.length !== vec_b.length) {
+            throw new Error("Vectors must have the same length");
+        }
+        let dot_product = 0;
+        let norm_a = 0;
+        let norm_b = 0;
+        for (let i = 0; i < vec_a.length; i++) {
+            dot_product += vec_a[i] * vec_b[i];
+            norm_a += vec_a[i] * vec_a[i];
+            norm_b += vec_b[i] * vec_b[i];
+        }
+        const denominator = Math.sqrt(norm_a) * Math.sqrt(norm_b);
+        if (denominator === 0) {
+            return 0;
+        }
+        return dot_product / denominator;
+    }
+    /**
+     * Optimized cosine similarity calculation that reuses pre-computed norms
+     * This avoids recalculating norms for every comparison
+     *
+     * @param vec_a - Query vector
+     * @param vec_b - Embedding vector to compare
+     * @param vec_a_norm - Pre-computed norm of vec_a (sqrt of sum of squares)
+     * @param vec_b_norm - Optional pre-computed norm of vec_b (if not provided, calculates it)
+     * @returns Cosine similarity score (0-1)
+     */
+    cosine_similarity_optimized(vec_a, vec_b, vec_a_norm, vec_b_norm) {
+        if (vec_a.length !== vec_b.length) {
+            throw new Error("Vectors must have the same length");
+        }
+        let dot_product = 0;
+        // Calculate dot product
+        for (let i = 0; i < vec_a.length; i++) {
+            dot_product += vec_a[i] * vec_b[i];
+        }
+        // Use pre-computed norm if provided, otherwise calculate it
+        const norm_b = vec_b_norm || Math.sqrt(vec_b.reduce((sum, val) => sum + val * val, 0));
+        const denominator = vec_a_norm * norm_b;
+        if (denominator === 0) {
+            return 0;
+        }
+        return dot_product / denominator;
+    }
+}
+exports.EmbeddingService = EmbeddingService;
+//# sourceMappingURL=embedding.service.js.map

package/dist/services/embedding.service.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"embedding.service.js","sourceRoot":"","sources":["../../src/services/embedding.service.ts"],"names":[],"mappings":";AAAA;;;;;;;GAOG;;;AAEH,yCAA4C;AAC5C,gDAA4D;AAC5D,sCAA0E;AAC1E,+BAAoC;AACpC,sCAAyC;AACzC,gEAAiE;AACjE,yCAA+B;AAE/B,2CAAmC;AAEnC,MAAM,MAAM,GAAG,IAAA,8BAAqB,EAAC,mBAAmB,CAAC,CAAC;AAE1D;;GAEG;AACH,SAAS,aAAa,CAAC,IAAY,EAAE,aAAqB,IAAI,EAAE,UAAkB,GAAG;IACjF,MAAM,MAAM,GAAa,EAAE,CAAC;IAC5B,IAAI,KAAK,GAAG,CAAC,CAAC;IAEd,OAAO,KAAK,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC;QACzB,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,GAAG,UAAU,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC;QACtD,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,CAAC;QACpC,IAAI,GAAG,IAAI,IAAI,CAAC,MAAM;YAAE,MAAM;QAC9B,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,KAAK,GAAG,CAAC,EAAE,GAAG,GAAG,OAAO,CAAC,CAAC;IAC/C,CAAC;IAED,OAAO,MAAM,CAAC;AAClB,CAAC;AAED,MAAa,gBAAgB;IAiBzB;QAfQ,kBAAa,GAAW,oBAAoB,CAAC,CAAC,yBAAyB;QAE/E,+CAA+C;QAC/C,qEAAqE;QACrE,8EAA8E;QACtE,0BAAqB,GAAG,IAAI,GAAG,EAA2B,CAAC;QAClD,2BAAsB,GAAG,EAAE,GAAG,EAAE,GAAG,IAAI,CAAC,CAAC,SAAS;QAClD,6BAAwB,GAAG,IAAI,CAAC;QAChC,kCAA6B,GAAG,IAAI,CAAC,CAAC,mDAAmD;QACzF,wBAAmB,GAAG,EAAE,CAAC,CAAC,yDAAyD;QAEpG,+CAA+C;QACvC,sBAAiB,GAAG,IAAI,GAAG,EAA0B,CAAC;QACtD,sBAAiB,GAAG,KAAK,CAAC;QAG9B,MAAM,OAAO,GAAG,qBAAY,CAAC,GAAG,CAAC,gBAAgB,CAAC,CAAC;QACnD,IAAI,CAAC,OAAO,EAAE,CAAC;YACX,MAAM,IAAI,KAAK,CAAC,+BAA+B,CAAC,CAAC;QACrD,CAAC;QACD,IAAI,CAAC,MAAM,GAAG,IAAI,mBAAW,CAAC,EAAE,MAAM,EAAE,OAAiB,EAAE,CAAC,CAAC;IACjE,CAAC;IAED;;;;OAIG;IACK,KAAK,CAAC,wBAAwB,CAAC,KAAa;QAChD,IAAI,IAAI,CAAC,iBAAiB,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,CAAC;YACpC,OAAO,IAAI,CAAC,iBAAiB,CAAC,GAAG,CAAC,KAAK,CAAE,CAAC;QAC9C,CAAC;QAED,IAAI,IAAI,CAAC,iBAAiB,EAAE,CAAC;YACzB,MAAM,CAAC,IAAI,CAAC,gDAAgD,CAAC,CAAC;YAC9D,uEAAuE;YACvE,MAAM,IAAI,OAAO,CAAC,OAAO,CAAC,EAAE,CAAC,UAAU,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC,CAAC;YACxD,OAAO,IAAI,CAAC,wBAAwB,CAAC,KAAK,CAAC,CAAC;QAChD,CAAC;QAED,IAAI,CAAC;YACD,IAAI,CAAC,iBAAiB,GAAG,IAAI,CAAC;YAC9B,MAAM,CAAC,IAAI,CAAC,mCAAmC,KAAK,KAAK,CAAC,CAAC;YAE3D,MAAM,UAAU,GAAG,MAAM,uBAAc,CAAC,OAAO,CAAC;gBAC5C,KAAK,EAAE,EAAE,KAAK,EAAE;gBAChB,UAAU,EAAE,CAAC,IAAI,EAAE,WAAW,EAAE,MAAM,EAAE,cAAc,EAAE,uBAAuB,EAAE,gBAAgB,EAAE,aAAa,EAAE,MAAM,EAAE,QAAQ,CAAC;aACtI,CAAC,CAAC;YAEH,IAAI,UAAU,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBAC1B,MAAM,CAAC,IAAI,CAAC,iCAAiC,KAAK,kBAAkB,CAAC,CAAC;gBACtE,OAAO,IAAI,CAAC;YAChB,CAAC;YAED,MAAM,SAAS,GAAG,IAAI,CAAC,KAAK,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,MAAM,CAAC;YAC7D,MAAM,KAAK,GAAG,IAAI,kBAAK,CAAC,SAAS,CAAC,CAAC;YAEnC,MAAM,OAAO,GAAe,EAAE,CAAC;YAC/B,MAAM,QAAQ,GAA+B,EAAE,CAAC;YAEhD,KAAK,MAAM,GAAG,IAAI,UAAU,EAAE,CAAC;gBAC3B,IAAI,CAAC;oBACD,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,SAAS,CAAC,CAAC;oBACzC,4BAA4B;oBAC5B,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC,IAAI,MAAM,CAAC,MAAM,KAAK,SAAS,EAAE,CAAC;wBACxD,MAAM,CAAC,IAAI,CAAC,8CAA8C,GAAG,CAAC,EAAE,cAAc,SAAS,SAAS,KAAK,CAAC,OAAO,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC,CAAC,cAAc,GAAG,CAAC,CAAC;wBAC3J,SAAS;oBACb,CAAC;oBACD,OAAO,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;oBACrB,QAAQ,CAAC,IAAI,CAAC;wBACV,EAAE,EAAE,GAAG,CAAC,EAAE;wBACV,IAAI,EAAE,GAAG,CAAC,IAAI;wBACd,YAAY,EAAE,GAAG,CAAC,YAAY;wBAC9B,qBAAqB,EAAE,GAAG,CAAC,qBAAqB;wBAChD,cAAc,EAAE,GAAG,CAAC,cAAc;wBAClC,WAAW,EAAE,GAAG,CAAC,WAAW;wBAC5B,IAAI,EAAE,GAAG,CAAC,IAAI;wBACd,MAAM,EAAE,GAAG,CAAC,MAAM;qBACrB,CAAC,CAAC;gBACP,CAAC;gBAAC,OAAO,CAAC,EAAE,CAAC;oBACT,MAAM,CAAC,IAAI,CAAC,sCAAsC,GAAG,CAAC,EAAE,EAAE,EAAE,EAAE,KAAK,EAAE,CAAC,YAAY,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;gBACvH,CAAC;YACL,CAAC;YAED,IAAI,OAAO,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;gBACrB,0EAA0E;gBAC1E,KAAK,MAAM,MAAM,IAAI,OAAO,EAAE,CAAC;oBAC3B,KAAK,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;gBACtB,CAAC;gBACD,MAAM,UAAU,GAAmB,EAAE,KAAK,EAAE,QAAQ,EAAE,CAAC;gBACvD,IAAI,CAAC,iBAAiB,CAAC,GAAG,CAAC,KAAK,EAAE,UAAU,CAAC,CAAC;gBAC9C,MAAM,CAAC,IAAI,CAAC,iCAAiC,KAAK,SAAS,KAAK,CAAC,MAAM,EAAE,WAAW,CAAC,CAAC;gBACtF,OAAO,UAAU,CAAC;YACtB,CAAC;YACD,OAAO,IAAI,CAAC;QAChB,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YACb,MAAM,CAAC,KAAK,CAAC,2CAA2C,KAAK,EAAE,EAAE,EAAE,KAAK,EAAE,CAAC,CAAC;YAC5E,OAAO,IAAI,CAAC;QAChB,CAAC;gBAAS,CAAC;YACP,IAAI,CAAC,iBAAiB,GAAG,KAAK,CAAC;QACnC,CAAC;IACL,CAAC;IAED;;;;;;;;;;;OAWG;IACK,KAAK,CAAC,yBAAyB,CAAC,IAAY,EAAE,KAAa;QAC/D,MAAM,eAAe,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC,IAAI,EAAE,CAAC;QAClD,MAAM,eAAe,GAAG,GAAG,KAAK,IAAI,eAAe,EAAE,CAAC;QAEtD,iFAAiF;QACjF,MAAM,YAAY,GAAG,IAAI,CAAC,qBAAqB,CAAC,GAAG,CAAC,eAAe,CAAC,CAAC;QACrE,IAAI,YAAY,IAAI,IAAI,CAAC,GAAG,EAAE,GAAG,YAAY,CAAC,SAAS,GAAG,IAAI,CAAC,sBAAsB,EAAE,CAAC;YACpF,MAAM,CAAC,KAAK,CAAC,yCAAyC,EAAE;gBACpD,SAAS,EAAE,eAAe,CAAC,SAAS,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,KAAK;gBACnD,UAAU,EAAE,IAAI,CAAC,qBAAqB,CAAC,IAAI;aAC9C,CAAC,CAAC;YACH,OAAO,YAAY,CAAC,SAAS,CAAC;QAClC,CAAC;QAED,yEAAyE;QACzE,MAAM,eAAe,GAAG,MAAM,IAAI,CAAC,kBAAkB,CAAC,IAAI,EAAE,KAAK,CAAC,CAAC;QACnE,MAAM,UAAU,GAAG,IAAI,CAAC,IAAI,CAAC,eAAe,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,GAAG,EAAE,EAAE,CAAC,GAAG,GAAG,GAAG,GAAG,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC;QAEvF,8EAA8E;QAC9E,kCAAkC;QAClC,IAAI,IAAI,CAAC,qBAAqB,CAAC,IAAI,GAAG,CAAC,EAAE,CAAC;YACtC,MAAM,aAAa,GAAG,IAAI,CAAC,kCAAkC,CACzD,eAAe,EACf,UAAU,EACV,KAAK,EACL,IAAI,CAAC,6BAA6B,CACrC,CAAC;YAEF,IAAI,aAAa,EAAE,CAAC;gBAChB,MAAM,CAAC,KAAK,CAAC,4CAA4C,EAAE;oBACvD,UAAU,EAAE,IAAI,CAAC,SAAS,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,KAAK;oBACzC,UAAU,EAAE,IAAI,CAAC,qBAAqB,CAAC,IAAI;oBAC3C,SAAS,EAAE,IAAI,CAAC,6BAA6B;iBAChD,CAAC,CAAC;gBACH,+DAA+D;gBAC/D,wDAAwD;gBACxD,IAAI,CAAC,eAAe,CAAC,eAAe,EAAE,eAAe,EAAE,UAAU,EAAE,IAAI,CAAC,CAAC;gBACzE,OAAO,aAAa,CAAC;YACzB,CAAC;QACL,CAAC;QAED,mDAAmD;QACnD,MAAM,CAAC,KAAK,CAAC,mDAAmD,EAAE;YAC9D,SAAS,EAAE,eAAe,CAAC,SAAS,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,KAAK;YACnD,UAAU,EAAE,IAAI,CAAC,qBAAqB,CAAC,IAAI;SAC9C,CAAC,CAAC;QACH,IAAI,CAAC,eAAe,CAAC,eAAe,EAAE,eAAe,EAAE,UAAU,EAAE,IAAI,CAAC,CAAC;QACzE,OAAO,eAAe,CAAC;IAC3B,CAAC;IAED;;;OAGG;IACK,eAAe,CACnB,SAAiB,EACjB,SAAmB,EACnB,cAAuB,EACvB,aAAsB;QAEtB,mCAAmC;QACnC,MAAM,IAAI,GAAG,cAAc,IAAI,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,GAAG,EAAE,EAAE,CAAC,GAAG,GAAG,GAAG,GAAG,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC;QAE7F,qDAAqD;QACrD,IAAI,IAAI,CAAC,qBAAqB,CAAC,IAAI,IAAI,IAAI,CAAC,wBAAwB,EAAE,CAAC;YACnE,MAAM,UAAU,GAAG,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,qBAAqB,CAAC,OAAO,EAAE,CAAC;iBAC9D,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,SAAS,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YAC3D,IAAI,CAAC,qBAAqB,CAAC,MAAM,CAAC,UAAU,CAAC,CAAC;YAC9C,MAAM,CAAC,KAAK,CAAC,4BAA4B,EAAE;gBACvC,WAAW,EAAE,UAAU,CAAC,SAAS,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,KAAK;gBAChD,cAAc,EAAE,IAAI,CAAC,qBAAqB,CAAC,IAAI;aAClD,CAAC,CAAC;QACP,CAAC;QAED,IAAI,CAAC,qBAAqB,CAAC,GAAG,CAAC,SAAS,EAAE;YACtC,SAAS;YACT,cAAc,EAAE,IAAI;YACpB,SAAS,EAAE,IAAI,CAAC,GAAG,EAAE;YACrB,aAAa,EAAE,aAAa,IAAI,SAAS;SAC5C,CAAC,CAAC;IACP,CAAC;IAED;;;;;;;;;;;;OAYG;IACK,kCAAkC,CACtC,eAAyB,EACzB,UAAkB,EAClB,KAAa,EACb,oBAA4B;QAE5B,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;QACvB,IAAI,aAAa,GAAG,CAAC,CAAC;QACtB,IAAI,UAAU,GAAuD,IAAI,CAAC;QAE1E,yDAAyD;QACzD,KAAK,MAAM,CAAC,SAAS,EAAE,MAAM,CAAC,IAAI,IAAI,CAAC,qBAAqB,CAAC,OAAO,EAAE,EAAE,CAAC;YACrE,4CAA4C;YAC5C,IAAI,CAAC,SAAS,CAAC,UAAU,CAAC,GAAG,KAAK,GAAG,CAAC,EAAE,CAAC;gBACrC,SAAS;YACb,CAAC;YAED,iCAAiC;YACjC,IAAI,GAAG,GAAG,MAAM,CAAC,SAAS,IAAI,IAAI,CAAC,sBAAsB,EAAE,CAAC;gBACxD,SAAS;YACb,CAAC;YAED,mEAAmE;YACnE,IAAI,aAAa,IAAI,IAAI,CAAC,mBAAmB,EAAE,CAAC;gBAC5C,MAAM;YACV,CAAC;YAED,gDAAgD;YAChD,MAAM,UAAU,GAAG,IAAI,CAAC,2BAA2B,CAC/C,eAAe,EACf,MAAM,CAAC,SAAS,EAChB,UAAU,EACV,MAAM,CAAC,cAAc,CACxB,CAAC;YAEF,aAAa,EAAE,CAAC;YAEhB,uFAAuF;YACvF,IAAI,CAAC,UAAU,IAAI,UAAU,GAAG,UAAU,CAAC,UAAU,EAAE,CAAC;gBACpD,UAAU,GAAG,EAAE,SAAS,EAAE,MAAM,CAAC,SAAS,EAAE,UAAU,EAAE,CAAC;YAC7D,CAAC;YAED,wDAAwD;YACxD,IAAI,UAAU,IAAI,oBAAoB,EAAE,CAAC;gBACrC,MAAM,CAAC,KAAK,CAAC,6CAA6C,EAAE;oBACxD,SAAS,EAAE,SAAS,CAAC,SAAS,CAAC,CAAC,EAAE,EAAE,CAAC,GAAG,KAAK;oBAC7C,UAAU,EAAE,UAAU,CAAC,OAAO,CAAC,CAAC,CAAC;oBACjC,aAAa;iBAChB,CAAC,CAAC;gBACH,OAAO,MAAM,CAAC,SAAS,CAAC;YAC5B,CAAC;QACL,CAAC;QAED,4CAA4C;QAC5C,IAAI,aAAa,GAAG,CAAC,IAAI,UAAU,EAAE,CAAC;YAClC,MAAM,CAAC,KAAK,CAAC,yCAAyC,EAAE;gBACpD,eAAe,EAAE,UAAU,CAAC,UAAU,CAAC,OAAO,CAAC,CAAC,CAAC;gBACjD,SAAS,EAAE,oBAAoB;gBAC/B,aAAa;gBACb,UAAU,EAAE,IAAI,CAAC,qBAAqB,CAAC,IAAI;aAC9C,CAAC,CAAC;QACP,CAAC;QAED,OAAO,IAAI,CAAC;IAChB,CAAC;IAED;;;;;;OAMG;IACH,KAAK,CAAC,kBAAkB,CAAC,IAAY,EAAE,KAAc;QACjD,IAAI,CAAC;YACD,MAAM,eAAe,GAAG,KAAK,IAAI,IAAI,CAAC,aAAa,CAAC;YAEpD,MAAM,CAAC,KAAK,CAAC,mCAAmC,EAAE;gBAC9C,KAAK,EAAE,eAAe;gBACtB,WAAW,EAAE,IAAI,CAAC,MAAM;aAC3B,CAAC,CAAC;YAEH,6BAA6B;YAC7B,4DAA4D;YAC5D,MAAM,MAAM,GAAG,MAAM,IAAI,CAAC,MAAM,CAAC,MAAM,CAAC,YAAY,CAAC;gBACjD,KAAK,EAAE,eAAe;gBACtB,QAAQ,EAAE,CAAC;wBACP,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC;qBACpB,CAAC;aACL,CAAC,CAAC;YAEH,6CAA6C;YAC7C,qDAAqD;YACrD,MAAM,gBAAgB,GAAG,MAAM,CAAC,UAAU,EAAE,CAAC,CAAC,CAAC,EAAE,MAAM,CAAC;YACxD,IAAI,CAAC,gBAAgB,IAAI,gBAAgB,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBACrD,MAAM,IAAI,KAAK,CAAC,mCAAmC,CAAC,CAAC;YACzD,CAAC;YAED,MAAM,CAAC,KAAK,CAAC,kCAAkC,EAAE;gBAC7C,KAAK,EAAE,eAAe;gBACtB,mBAAmB,EAAE,gBAAgB,CAAC,MAAM;aAC/C,CAAC,CAAC;YAEH,OAAO,gBAAgB,CAAC;QAC5B,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YACb,MAAM,CAAC,KAAK,CAAC,8BAA8B,EAAE,EAAE,KAAK,EAAE,CAAC,CAAC;YACxD,MAAM,IAAI,KAAK,CAAC,iCAAiC,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,eAAe,EAAE,CAAC,CAAC;QACjH,CAAC;IACL,CAAC;IAED;;;;;;;;;;;;;;OAcG;IACI,KAAK,CAAC,8BAA8B,CACvC,YAAoB,EACpB,qBAA6B,EAC7B,IAAY,EACZ,cAAkC,EAClC,KAAyB,EACzB,IAAmB,EACnB,MAAc;QAEd,+BAA+B;QAC/B,IAAI,CAAC,YAAY,IAAI,CAAC,qBAAqB,EAAE,CAAC;YAC1C,MAAM,IAAI,KAAK,CAAC,6EAA6E,CAAC,CAAC;QACnG,CAAC;QAED,kBAAkB;QAClB,IAAI,CAAC,MAAM,IAAI,MAAM,CAAC,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YACxC,MAAM,IAAI,KAAK,CAAC,gEAAgE,CAAC,CAAC;QACtF,CAAC;QAED,gDAAgD;QAChD,MAAM,iBAAiB,GAAG,MAAM,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;QACtD,MAAM,eAAe,GAAG,KAAK,IAAI,IAAI,CAAC,aAAa,CAAC;QAEpD,MAAM,CAAC,IAAI,CAAC,oCAAoC,EAAE;YAC9C,YAAY;YACZ,qBAAqB;YACrB,YAAY,EAAE,CAAC,EAAE,6BAA6B;YAC9C,KAAK,EAAE,eAAe;YACtB,MAAM,EAAE,iBAAiB;SAC5B,CAAC,CAAC;QAEH,iBAAiB;QACjB,MAAM,MAAM,GAAG,aAAa,CAAC,IAAI,EAAE,IAAI,EAAE,GAAG,CAAC,CAAC;QAE9C,MAAM,CAAC,IAAI,CAAC,kBAAkB,EAAE;YAC5B,YAAY;YACZ,YAAY,EAAE,MAAM,CAAC,MAAM;YAC3B,MAAM,EAAE,iBAAiB;SAC5B,CAAC,CAAC;QAEH,8DAA8D;QAC9D,wEAAwE;QACpE,MAAM,qBAAqB,GAAU,EAAE,CAAC;QACxC,MAAM,UAAU,GAAG,EAAE,CAAC;QAEtB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,CAAC,MAAM,EAAE,CAAC,IAAI,UAAU,EAAE,CAAC;YACjD,MAAM,KAAK,GAAG,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,GAAG,UAAU,CAAC,CAAC;YAElD,uEAAuE;YACnE,MAAM,kBAAkB,GAAG,KAAK,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,KAAK,EAAE,EAAE,CAClD,IAAI,CAAC,kBAAkB,CAAC,KAAK,EAAE,eAAe,CAAC,CAAC,IAAI,CAAC,CAAC,SAAS,EAAE,EAAE,CAAC,CAAC;gBACjE,WAAW,EAAE,CAAC,GAAG,KAAK;gBACtB,IAAI,EAAE,KAAK;gBACX,SAAS;aACZ,CAAC,CAAC,CACN,CAAC;YAEF,MAAM,OAAO,GAAG,MAAM,OAAO,CAAC,GAAG,CAAC,kBAAkB,CAAC,CAAC;YAEtD,6CAA6C;YAC7C,KAAK,MAAM,MAAM,IAAI,OAAO,EAAE,CAAC;gBAC/B,MAAM,gBAAgB,GAAQ;oBACtB,EAAE,EAAE,IAAA,SAAM,GAAE;oBACZ,YAAY;oBACZ,qBAAqB;oBACrB,cAAc;oBACd,IAAI,EAAE,MAAM,CAAC,IAAI;oBACjB,SAAS,EAAE,IAAI,CAAC,SAAS,CAAC,MAAM,CAAC,SAAS,CAAC;oBAC3C,KAAK,EAAE,eAAe;oBACtB,WAAW,EAAE,MAAM,CAAC,WAAW;oBAC/B,IAAI,EAAE,IAAI,EAAE,4CAA4C;iBAC/D,CAAC;gBAEF,8DAA8D;gBAC9D,oDAAoD;gBACpD,gBAAgB,CAAC,MAAM,GAAG,iBAAiB,CAAC;gBAE5C,qBAAqB,CAAC,IAAI,CAAC,gBAAgB,CAAC,CAAC;YAC7C,CAAC;YAEL,MAAM,CAAC,KAAK,CAAC,4BAA4B,EAAE;gBACvC,aAAa,EAAE,qBAAqB,CAAC,MAAM;gBAC3C,MAAM;aACT,CAAC,CAAC;QACP,CAAC;QAED,yDAAyD;QACzD,uEAAuE;QACvE,MAAM,SAAS,GAAG,IAAA,yCAAqB,GAAE,CAAC;QAC1C,MAAM,WAAW,GAAG,MAAM,SAAS,CAAC,WAAW,EAAE,CAAC;QAElD,IAAI,CAAC;YACD,IAAI,kBAAkB,GAAG,CAAC,CAAC;YAE3B,uEAAuE;YACvE,IAAI,qBAAqB,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;gBACnC,MAAM,uBAAc,CAAC,UAAU,CAAC,qBAAqB,EAAE,EAAE,WAAW,EAAE,CAAC,CAAC;gBACxE,kBAAkB,GAAG,qBAAqB,CAAC,MAAM,CAAC;YACtD,CAAC;YAED,MAAM,WAAW,CAAC,MAAM,EAAE,CAAC;YAE3B,MAAM,CAAC,IAAI,CAAC,iCAAiC,EAAE;gBAC3C,YAAY;gBACZ,qBAAqB;gBACrB,kBAAkB;aACrB,CAAC,CAAC;YAEH,OAAO,EAAE,kBAAkB,EAAE,YAAY,EAAE,qBAAqB,EAAE,CAAC;QACvE,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YACb,MAAM,WAAW,CAAC,QAAQ,EAAE,CAAC;YAC7B,MAAM,CAAC,KAAK,CAAC,4BAA4B,EAAE,EAAE,YAAY,EAAE,qBAAqB,EAAE,KAAK,EAAE,CAAC,CAAC;YAC3F,MAAM,KAAK,CAAC;QAChB,CAAC;IACL,CAAC;IAED;;OAEG;IACI,KAAK,CAAC,kCAAkC;QAC3C,IAAI,CAAC;YAED,MAAM,MAAM,GAAG,qBAAY,CAAC,GAAG,CAAC,gBAAgB,CAAC,IAAI,sBAAsB,CAAC;YAC5E,MAAM,YAAY,GAAG,qBAAY,CAAC,GAAG,CAAC,cAAc,CAAC,IAAI,mBAAmB,CAAC;YAE7E,MAAM,iBAAiB,GAAG,MAAM,uBAAc,CAAC,OAAO,CAAC;gBACnD,KAAK,EAAE;oBACH,cAAc,EAAE,MAAgB;oBAChC,YAAY,EAAE,YAAsB;oBACpC,QAAQ,EAAE;wBACN,CAAC,cAAE,CAAC,IAAI,CAAC,EAAE,8BAA8B;qBAC5C;iBACJ;gBACD,KAAK,EAAE,CAAC;gBACR,KAAK,EAAE,CAAC,CAAC,YAAY,EAAE,MAAM,CAAC,CAAC;aAClC,CAAC,CAAC;YAEH,OAAO,iBAAiB,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,iBAAiB,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC;QACzE,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YACb,MAAM,CAAC,KAAK,CAAC,oCAAoC,EAAE,EAAE,KAAK,EAAE,CAAC,CAAC;YAC9D,OAAO,IAAI,CAAC;QAChB,CAAC;IACL,CAAC;IAED;;;;;;;;;;OAUG;IACH,KAAK,CAAC,iBAAiB,CACnB,UAAkB,EAClB,YAAqB,EACrB,QAAgB,EAAE,EAClB,KAAc,EACd,IAAoB,EACpB,MAAe;QAEf,IAAI,CAAC;YACD,MAAM,eAAe,GAAG,KAAK,IAAI,IAAI,CAAC,aAAa,CAAC;YAEpD,8EAA8E;YAC9E,0DAA0D;YAC1D,IAAI,qBAAqB,GAAuB,IAAI,CAAC;YACrD,IAAI,YAAY,IAAI,IAAI,IAAI,MAAM,EAAE,CAAC;gBACjC,MAAM,YAAY,GAAQ,EAAE,KAAK,EAAE,eAAe,EAAE,CAAC;gBAErD,IAAI,YAAY,EAAE,CAAC;oBACf,YAAY,CAAC,YAAY,GAAG,YAAY,CAAC;gBAC7C,CAAC;gBAED,IAAI,IAAI,EAAE,CAAC;oBACP,YAAY,CAAC,IAAI,GAAG,IAAI,CAAC;gBAC7B,CAAC;gBAED,IAAI,MAAM,EAAE,CAAC;oBACT,iEAAiE;oBACjE,YAAY,CAAC,MAAM,GAAG,MAAM,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;gBACtD,CAAC;gBAED,MAAM,mBAAmB,GAAG,MAAM,uBAAc,CAAC,OAAO,CAAC;oBACrD,KAAK,EAAE,YAAY;oBACnB,UAAU,EAAE,CAAC,IAAI,CAAC;iBACzB,CAAC,CAAC;gBAEC,IAAI,mBAAmB,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;oBACnC,MAAM,CAAC,KAAK,CAAC,sCAAsC,EAAE,EAAE,YAAY,EAAE,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC;oBACrF,OAAO,EAAE,CAAC;gBAClB,CAAC;gBAEG,qBAAqB,GAAG,IAAI,GAAG,CAAC,mBAAmB,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC;gBACpE,MAAM,CAAC,KAAK,CAAC,iCAAiC,EAAE;oBAC5C,cAAc,EAAE,qBAAqB,CAAC,IAAI;oBAC1C,YAAY;oBACZ,IAAI;oBACJ,MAAM;iBACT,CAAC,CAAC;YACP,CAAC;YAED,qDAAqD;YACrD,MAAM,UAAU,GAAG,MAAM,IAAI,CAAC,wBAAwB,CAAC,eAAe,CAAC,CAAC;YACxE,IAAI,CAAC,UAAU,EAAE,CAAC;gBACd,MAAM,CAAC,IAAI,CAAC,uCAAuC,eAAe,0BAA0B,CAAC,CAAC;gBAC9F,OAAO,EAAE,CAAC;YACd,CAAC;YAED,sDAAsD;YACtD,MAAM,eAAe,GAAG,MAAM,IAAI,CAAC,yBAAyB,CAAC,UAAU,EAAE,eAAe,CAAC,CAAC;YAE1F,2BAA2B;YAC3B,MAAM,6BAA6B,GAAG,MAAM,IAAI,CAAC,kCAAkC,EAAE,CAAC;YAEtF,6CAA6C;YAC7C,wEAAwE;YACxE,sEAAsE;YACtE,MAAM,YAAY,GAAG,UAAU,CAAC,KAAK,CAAC,MAAM,EAAE,CAAC;YAC/C,MAAM,eAAe,GAAG,qBAAqB,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC;YACtE,MAAM,YAAY,GAAG,IAAI,CAAC,GAAG,CAAC,eAAe,EAAE,YAAY,CAAC,CAAC;YAE7D,IAAI,YAAY,KAAK,CAAC,EAAE,CAAC;gBACrB,MAAM,CAAC,KAAK,CAAC,yCAAyC,CAAC,CAAC;gBACxD,OAAO,EAAE,CAAC;YACd,CAAC;YAED,MAAM,EAAE,MAAM,EAAE,SAAS,EAAE,GAAG,UAAU,CAAC,KAAK,CAAC,MAAM,CAAC,eAAe,EAAE,YAAY,CAAC,CAAC;YAErF,sFAAsF;YACtF,MAAM,WAAW,GAAG,MAAM;iBACrB,GAAG,CAAC,CAAC,KAAa,EAAE,CAAS,EAAE,EAAE;gBAC9B,MAAM,QAAQ,GAAG,UAAU,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC;gBAC5C,0EAA0E;gBAC1E,MAAM,KAAK,GAAG,CAAC,GAAG,SAAS,CAAC,CAAC,CAAC,CAAC;gBAC/B,OAAO,EAAE,GAAG,QAAQ,EAAE,KAAK,EAAE,CAAC;YAClC,CAAC,CAAC;iBACD,MAAM,CAAC,CAAC,MAAW,EAAE,EAAE;gBACpB,yFAAyF;gBACzF,IAAI,qBAAqB,IAAI,CAAC,qBAAqB,CAAC,GAAG,CAAC,MAAM,CAAC,EAAE,CAAC,EAAE,CAAC;oBACjE,OAAO,KAAK,CAAC;gBACjB,CAAC;gBACD,OAAO,IAAI,CAAC;YAChB,CAAC,CAAC,CAAC;YAEP,iCAAiC;YACjC,MAAM,aAAa,GAAG,WAAW;iBAC5B,IAAI,CAAC,CAAC,CAAM,EAAE,CAAM,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC;iBAC3C,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC;YAErB,IAAI,aAAa,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBAC7B,MAAM,CAAC,IAAI,CAAC,4CAA4C,EAAE,EAAE,YAAY,EAAE,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC;gBAC1F,OAAO,EAAE,CAAC;YACd,CAAC;YAED,MAAM,CAAC,KAAK,CAAC,8BAA8B,EAAE;gBACzC,aAAa,EAAE,MAAM,CAAC,MAAM;gBAC5B,cAAc,EAAE,WAAW,CAAC,MAAM;gBAClC,WAAW,EAAE,aAAa,CAAC,MAAM;gBACjC,KAAK;gBACL,WAAW,EAAE,CAAC,CAAC,qBAAqB;aACvC,CAAC,CAAC;YAEH,OAAO,aAAa,CAAC;QACzB,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YACb,MAAM,CAAC,KAAK,CAAC,6BAA6B,EAAE,EAAE,KAAK,EAAE,CAAC,CAAC;YACvD,MAAM,KAAK,CAAC;QAChB,CAAC;IACL,CAAC;IAED;;;;;;OAMG;IACI,iBAAiB,CAAC,KAAe,EAAE,KAAe;QACrD,IAAI,KAAK,CAAC,MAAM,KAAK,KAAK,CAAC,MAAM,EAAE,CAAC;YAChC,MAAM,IAAI,KAAK,CAAC,mCAAmC,CAAC,CAAC;QACzD,CAAC;QAED,IAAI,WAAW,GAAG,CAAC,CAAC;QACpB,IAAI,MAAM,GAAG,CAAC,CAAC;QACf,IAAI,MAAM,GAAG,CAAC,CAAC;QAEf,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACpC,WAAW,IAAI,KAAK,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;YACnC,MAAM,IAAI,KAAK,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;YAC9B,MAAM,IAAI,KAAK,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QAClC,CAAC;QAED,MAAM,WAAW,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAC1D,IAAI,WAAW,KAAK,CAAC,EAAE,CAAC;YACpB,OAAO,CAAC,CAAC;QACb,CAAC;QAED,OAAO,WAAW,GAAG,WAAW,CAAC;IACrC,CAAC;IAED;;;;;;;;;OASG;IACK,2BAA2B,CAC/B,KAAe,EACf,KAAe,EACf,UAAkB,EAClB,UAAmB;QAEnB,IAAI,KAAK,CAAC,MAAM,KAAK,KAAK,CAAC,MAAM,EAAE,CAAC;YAChC,MAAM,IAAI,KAAK,CAAC,mCAAmC,CAAC,CAAC;QACzD,CAAC;QAED,IAAI,WAAW,GAAG,CAAC,CAAC;QAEpB,wBAAwB;QACxB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE,CAAC;YACpC,WAAW,IAAI,KAAK,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QACvC,CAAC;QAED,4DAA4D;QAC5D,MAAM,MAAM,GAAG,UAAU,IAAI,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,GAAG,EAAE,EAAE,CAAC,GAAG,GAAG,GAAG,GAAG,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC;QACvF,MAAM,WAAW,GAAG,UAAU,GAAG,MAAM,CAAC;QAExC,IAAI,WAAW,KAAK,CAAC,EAAE,CAAC;YACpB,OAAO,CAAC,CAAC;QACb,CAAC;QAED,OAAO,WAAW,GAAG,WAAW,CAAC;IACrC,CAAC;CACJ;AApqBD,4CAoqBC"}

package/dist/services/file_upload.service.d.ts

@@ -0,0 +1,50 @@
+/**
+ * @fileoverview File Upload Service
+ *
+ * Service for handling file uploads and storing metadata in source_data table.
+ *
+ * @author MeasureOne, Inc.
+ * @version 1.0.0
+ */
+import { SourceDataModel } from "../models";
+export declare class FileUploadService {
+    private uploads_dir;
+    constructor();
+    /**
+     * Ensure uploads directory exists
+     */
+    private ensure_uploads_directory;
+    /**
+     * Save uploaded file and create source_data record
+     *
+     * @param file - Multer file object
+     * @param application_id - Optional application ID
+     * @param requester_id - Optional requester ID
+     * @returns Promise resolving to file_id
+     */
+    save_file(file: Express.Multer.File, application_id?: string, requester_id?: string): Promise<{
+        file_id: string;
+    }>;
+    /**
+     * Get file path by file_id
+     *
+     * @param file_id - Source data ID
+     * @returns Promise resolving to file path
+     */
+    get_file_path(file_id: string): Promise<string>;
+    /**
+     * Read file content by file_id
+     *
+     * @param file_id - Source data ID
+     * @returns Promise resolving to file content as string
+     */
+    read_file_content(file_id: string): Promise<string>;
+    /**
+     * Get source data by ID
+     *
+     * @param file_id - Source data ID
+     * @returns Promise resolving to source data
+     */
+    get_source_data(file_id: string): Promise<SourceDataModel>;
+}
+//# sourceMappingURL=file_upload.service.d.ts.map